Rethinking the Colonoscopy: The 13-Year NordICC Trial Verdict
It is one of the most dreaded, invasive, and universally recommended cancer screening procedures. But does it actually save lives? A landmark paper published in The Lancet challenges the assumed mortality benefits of population-wide colonoscopy screening.
1. The Invasive Standard of Care
For decades, the colonoscopy has been heralded as the gold standard for colorectal cancer (CRC) screening. The logic is mechanistically sound: insert a camera into the colon, find pre-cancerous polyps, and remove them before they can metastasize. However, the procedure requires a highly unpleasant bowel preparation regimen, sedation, and time off work. Despite its widespread implementation, particularly in the United States, colonoscopy screening was originally introduced without high-quality evidence from randomized controlled trials.
To fill this critical evidence gap, the Northern-European Initiative on Colorectal Cancer (NordICC) trial was initiated. This multicountry, population-based randomized controlled trial included 84,583 men and women aged 55-64 years from Norway, Poland, and Sweden. Participants were randomly allocated in a 1:2 ratio to either receive an invitation to undergo a single screening colonoscopy or to receive no screening. Now, with 13 years of follow-up data published in The Lancet by Kaminski et al. (2026), the results provide longer-term randomized evidence about the effects of an invitation to colonoscopy screening on colorectal cancer incidence and mortality.
2. The 13-Year Data: Incidence Down, Mortality Unchanged
The primary outcomes of the NordICC trial evaluated after 13 years were CRC incidence and mortality, analyzed strictly on an intention-to-screen basis (which reflects real-world public health outcomes).
Did the screening prevent cancer? Yes. The 13-year risk of developing CRC was 1.46% (375 of 28,217) in the screening invitation group, compared to 1.80% (912 of 56,366) in the no-screening group. This translates to a risk ratio (RR) of 0.81 (95% CI 0.71–0.90), meaning being invited to a colonoscopy reduced the incidence of colorectal cancer by roughly 19%.
However, the crucial metric is whether the procedure prevented death. The mortality result was inconclusive. The 13-year CRC mortality was 0.41% (106 deaths) in the screening group versus 0.47% (236 deaths) in the no-screening group. This yielded an intention-to-screen RR of 0.88 (95% CI 0.68–1.08), which was not statistically significant. The trial therefore did not demonstrate a statistically significant reduction in CRC mortality over 13 years; the confidence interval remains compatible with both a meaningful reduction and little or no reduction. This is not proof that colonoscopy has no mortality benefit, and longer follow-up may provide additional information.
A useful way to read these figures is to separate relative and absolute effects. The difference in observed 13-year CRC incidence was 1.80% − 1.46% = 0.34 percentage points, or about 3.4 fewer diagnoses per 1,000 people invited. The reciprocal of this trial-level risk difference is roughly 294 invitations per one fewer CRC diagnosis over 13 years. This is a descriptive translation of the intention-to-screen result—not a number needed to colonoscope, a prediction for an individual, or evidence of a mortality benefit—and it depends on the trial population, follow-up, and participation rate.
3. The Participation Problem and the "Per-Protocol" Reality
A major caveat of the NordICC trial is the human element: nobody likes getting a colonoscopy. Overall, only 42.0% (11,841 of 28,217) of the participants invited to the screening actually underwent the procedure.
When the researchers conducted a "per-protocol" analysis—estimating the effect if 100% of the invited individuals had actually been screened—the numbers improved. In the per-protocol analysis, CRC incidence dropped to 1.00% (RR 0.55; 95% CI 0.33–0.81), representing a 45% relative risk reduction. Yet, even among those who actually got the procedure, the per-protocol effect on CRC mortality remained highly imprecise (RR 0.70; 95% CI 0.26–1.25).
4. The Paradox of Modern Medicine
Why was a mortality benefit not demonstrated? When the NordICC trial was designed, CRC mortality in the no-screening group was projected to be 0.82%; the observed risk was 0.47%. The trial authors noted that mortality was lower than anticipated, which may have limited the ability to detect a difference between groups. The study does not establish why mortality was lower than expected, so improvements in treatment or other changes in population risk should be considered possible explanations rather than proven causes. The appropriate conclusion is narrower: this trial demonstrated a reduction in CRC incidence after an invitation to one-time colonoscopy, while its 13-year intention-to-screen analysis did not establish a statistically significant reduction in CRC mortality.
💡 My Practical Perspective: Is the Burden Justified?
For some individuals, bowel preparation, sedation, time away from work, and the invasive nature of colonoscopy are meaningful burdens. These practical costs should be considered alongside potential benefits, risks, and the likelihood that an invited person will complete screening. We willingly endure this burden under the promise that it is a life-saving necessity.
However, the 13-year NordICC data forces us to step back and weigh this heavy intervention rationally. In this trial, the absolute difference in CRC incidence was 0.34 percentage points over 13 years, while the mortality estimate remained uncertain. These results support careful comparison of screening strategies, including participation, harms, resources, and long-term outcomes; they do not by themselves establish that population-wide colonoscopy should be abandoned.
FIT is a non-invasive screening option used in established screening programs. Blood-based tests are also available for selected average-risk adults who need an alternative, but they are not interchangeable with colonoscopy: a positive result generally requires follow-up colonoscopy, and current guidance and test-specific evidence should be checked before describing any blood-based test as a first-line replacement.